Masan - Cuộc chiến Vonfram và thiên thời của tỷ phú Nguyễn Đăng Quang.

PHÂN TÍCH TỔNG HỢP
KIỂM SOÁT XUẤT KHẨU TUNGSTEN CỦA TRUNG QUỐC:
NGUYÊN NHÂN VÀ XÁC SUẤT CÁC KỊCH BẢN 5 NĂM TỚI (2026–2031)
Tổng hợp đồng thuận từ 6 hệ thống AI: ChatGPT • Claude • Copilot • DeepSeek • Gemini • Grok
Ngày: 14/04/2026

PHẦN I: TỔNG QUAN
Bối cảnh: Ngày 4/2/2025, Bộ Thương mại Trung Quốc (MOFCOM) ban hành kiểm soát xuất khẩu tungsten cùng tellurium, bismuth, molybdenum và indium theo Luật Kiểm soát Xuất khẩu và quy định về hàng lưỡng dụng. Kết quả ngay lập tức: xuất khẩu APT giảm ~70% (từ 782 tấn năm 2024 xuống 243 tấn trong 11 tháng đầu 2025), giá APT tăng 557%. Cho giai đoạn 2026–2027, chỉ 15 công ty được cấp phép xuất khẩu theo hệ thống whitelist.

Sáu hệ thống AI (ChatGPT, Claude, Copilot, DeepSeek, Gemini, Grok) đã được yêu cầu phân tích độc lập về (1) nguyên nhân Trung Quốc cấm xuất khẩu tungsten, và (2) xác suất các kịch bản trong 5 năm tới. Tài liệu này tổng hợp điểm chung và điểm khác biệt giữa các phân tích để xây dựng một bức tranh consensus đáng tin cậy.


PHẦN II: NGUYÊN NHÂN TRUNG QUỐC KIỂM SOÁT XUẤT KHẨU TUNGSTEN
Cả 6 hệ thống AI đều xác định được nhiều lớp động lực chồng chất, không chỉ đơn thuần là địa chính trị. Dưới đây là 5 nguyên nhân chính được đồng thuận:

1. Vũ khí địa chính trị – Đòn trả đũa trong chiến tranh thương mại Mỹ–Trung
Cả 6/6 AI đồng thuận: kiểm soát xuất khẩu tungsten là đòn phản công bất đối xứng (asymmetric retaliation) tương tự như gallium, germanium (2023) và antimony (2024). Lệnh kiểm soát được ban hành ngay sau khi Mỹ áp thuế bổ sung 10% lên hàng hóa Trung Quốc. Tungsten là vật liệu lưỡng dụng (dual-use): đạn xuyên giáp, hợp kim hàng không, bán dẫn, dụng cụ cắt CNC.

Grok và Claude nhấn mạnh thêm: chiến lược “tay thay chân” – Trung Quốc nhượng bộ về đất hiếm (rare earths deal 10/2025) nhưng đồng thời siết tungsten, antimony, bạc – duy trì đòn bẩy tổng thể trong khi tạo ảo giác “nhượng bộ”.

2. Chiến lược chuỗi giá trị – Từ “bán nguyên liệu” sang “bán sản phẩm”
5/6 AI (ChatGPT, Claude, Copilot, Gemini, Grok) xác định: Trung Quốc không muốn là “mỏ nguyên liệu” của thế giới. Bằng cách ngừng xuất khẩu APT, họ ép thế giới phải mua sản phẩm hạ nguồn có giá trị gia tăng cao (hợp kim cứng, dụng cụ cắt, linh kiện bán dẫn) từ Trung Quốc. Đây là chiến lược tương tự đã áp dụng với đất hiếm (2010s) và graphite/gallium (2023–2025).

ChatGPT mô tả sắc nét nhất: Trung Quốc đang chuyển từ “market supplier” sang “gatekeeper” – từ “ai có tiền mua được” sang “ai được duyệt mới mua được”. Đây là mô hình OPEC nhưng với một quốc gia duy nhất.

3. Cạn kiệt tài nguyên nội địa – Yếu tố cấu trúc bất khả nghịch
5/6 AI (trừ DeepSeek) nhấn mạnh: hàm lượng quặng tungsten Trung Quốc giảm từ 0,42% (2004) xuống 0,28% (2024). Hạn ngạch khai thác giảm 6,5% năm 2025 (xuống 58.000 tấn). Trung Quốc đã chuyển sang nhập khẩu ròng nguyên liệu thô (12.430 tấn năm 2024, tăng gấp đôi so với 2023).

Claude nhấn mạnh: đây là yếu tố bất khả nghịch – “bạn không thể bỏ cấm việc quặng hết.” Ngay cả khi địa chính trị hạ nhiệt, yếu tố vật lý này vẫn giữ nguồn cung thắt chặt.

4. Nhu cầu nội địa bùng nổ – Cạnh tranh trực tiếp với xuất khẩu
4/6 AI (ChatGPT, Claude, Grok, DeepSeek) xác định: nhu cầu nội địa APT của Trung Quốc đạt 110.000–120.000 tấn/năm, trong khi hạn ngạch khai thác chỉ 58.000 tấn. Các nguồn cầu mới bao gồm: dây cắt quang điện tungsten (tỷ lệ thâm nhập 20% → 40%), phụ gia cathode pin xe điện (+22% YoY), quốc phòng (+42% YoY), và tiềm năng nuclear fusion.
Điểm mấu chốt: xuất khẩu đang cạnh tranh trực tiếp với tiêu dùng nội địa. Đây là động lực kinh tế thuần túy, không liên quan địa chính trị.

5. Thể chế hóa hệ thống kiểm soát – Không còn là biện pháp tạm thời
3/6 AI (Claude, ChatGPT, Grok) nhấn mạnh: hệ thống whitelist 15 công ty cho 2026–2027 cho thấy đây đã trở thành kiến trúc quản lý dài hạn, không phải phản ứng nhất thời. Mọi lô hàng phải xin duyệt riêng (case-by-case), thời gian xử lý 30–60 ngày hoặc hơn, tạo ra sự không chắc chắn hành chính lan tỏa khắp chuỗi cung ứng.





PHẦN IV: ĐIỂM KHÁC BIỆT ĐÁNG CHÚ Ý GIỮA CÁC AI
ChatGPT – Lạc quan nhất về nới lỏng
ChatGPT là AI duy nhất đánh giá xác suất nới lỏng (có kiểm soát) lên đến 50–60%, cao hơn hẳn các AI khác. Tuy nhiên, ChatGPT nhấn mạnh rằng đây là “controlled reopening” chứ không phải free market return – giá vẫn sẽ cao.

DeepSeek – Bi quan nhất
DeepSeek đánh giá xác suất giữ nguyên (siết chặt có kiểm soát) lên đến 60%, siết chặt hơn 30%, và nới lỏng chỉ 10%. DeepSeek cũng là AI nhấn mạnh mạnh nhất yếu tố “an ninh quốc gia quan trọng hơn lợi nhuận” – dù giá tăng 5–6 lần, TQ vẫn không bán ra.Claude – Có framework phân tích sâu nhất
Claude là AI duy nhất tách riêng 4 kịch bản (bao gồm “mở hoàn toàn <2%”) và phân tích bằng framework “3 chân đỡ” độc lập (địa chính trị + cung vật lý + nhu cầu cấu trúc). Claude cũng dẫn nguồn “Nghịch lý cưỡng ép” (Pokorny 2026) và tiền lệ đất hiếm 2010.

Gemini & Copilot – Đơn giản nhưng hợp lý
Cả hai đều dùng 3 kịch bản cơ bản và đưa ra xác suất tương đồng. Gemini nổi bật với lập luận về môi trường (ô nhiễm nước từ tinh luyện APT tại Giang Tây), Copilot tập trung vào động lực “ngoại tệ vs chiến lược”.

PHẦN V: KẾT LUẬN VÀ HÀM Ý ĐẦU TƯ
Kết luận chính
• Nguyên nhân đa lớp: Địa chính trị là trigger nhưng không phải nguyên nhân duy nhất. Cạn kiệt tài nguyên, nhu cầu nội địa bùng nổ, và chiến lược chuỗi giá trị tạo nên “ba chân đỡ” giữ kiểm soát ổn định.
• Xác suất duy trì kiểm soát chặt: ~75–85%. Cả 6 AI đồng thuận: xác suất mở hoàn toàn dưới 10%. Kịch bản cơ sở là giữ nguyên hoặc siết chặt hơn.
• Thesis tungsten supercycle được củng cố: Ngay cả trong kịch bản nới lỏng (15–20%), giá APT vẫn sẽ cao hơn đáng kể so với pre-2025 do các yếu tố cấu trúc.
• Không phải single-variable bet: Rủi ro downside thực sự chỉ xảy ra khi cả 3 điều kiện hội tụ: TQ nới lỏng + nhu cầu sụp đổ + ex-China supply bùng nổ đồng thời – xác suất cực thấp.

Hàm ý cho giá APT
• Floor mới: Giá APT sẽ có mặt bằng mới cao hơn nhiều so với lịch sử, do xuất khẩu không còn elastic.
• Volatility cao: Mỗi lần cấp phép → giá giảm nhẹ; mỗi lần siết → giá spike. Biên độ dao động sẽ lớn.
• Phân vùng giá: Giá Trung Quốc (FOB) ≠ giá châu Âu (CIF Rotterdam). Phần bù địa chính trị (political premium) sẽ duy trì.
• Dài hạn (sau 2028–2030): Diversification (Mỹ–EU–Việt Nam–Úc–Canada) sẽ giảm dần phụ thuộc TQ, làm giá ổn định hơn ở mức cao mới.

Hàm ý cho MSR/MSN thesis
Phân tích tổng hợp này tăng cường đáng kể thesis tungsten supercycle và MSR/MSN deleverage-to-rerating:
• MSR (Masan High-Tech Materials) là nhà sản xuất tungsten lớn nhất ngoài Trung Quốc tại Việt Nam – hưởng lợi trực tiếp từ môi trường giá cao và nhu cầu diversification của phương Tây.
• Với xác suất 75–85% Trung Quốc duy trì kiểm soát, môi trường giá cao sẽ kéo dài đủ lâu để MSR chuyển từ lỗ sang lãi và MSN hoàn thành chu kỳ deleverage.
• Rủi ro chính không phải là “TQ nới lỏng” mà là thời gian ramp-up của các dự án ex-China cạnh tranh (Sangdong/Almonty, Mt Carbine/EQR). Tuy nhiên, các dự án này cần 3–5 năm để scale, tạo cửa sổ lợi nhuận đáng kể cho MSR trong giai đoạn 2026–2028.

Bảng giá đang trả lời kìa, các tin khác đều là bánh vẽ. Kiễn nhẫn chờ thời mã này xem 21 triệu trao tay giá nào? NĐT sẵn sàng khi tin tức đã biết trả giá nào ?

TT 2 phiên tăng 30đ, mà Q mắm nó vẫn giảm, hihi

Msr cầu mua trong phiên dưới 47 bị phang cho vỡ đầu . Nhà cái muốn nhả bớt phân phối vùng 45-47 cho nhỏ lẻ. Big boys muốn kéo bị phang cho vỡ mặt

Danny Lê mua MSN 4tr / 5tr đký nhé

Kể ra thằng tgđ này cũng còn chơi đẹp hơn Q mắm

2023 2024 MSR lỗ hơn 3k tỷ

--- Gộp bài viết, 14/04/2026 lúc 22:30, Bài cũ: 14/04/2026 lúc 22:28 ---

21tr cp gì vậy bạn?

MSR: 17/4 bắt đầu tháo cống

Mua trên sàn luôn. Uy tín đấy!

The Tungsten Throttle
How One Country Quietly Strangled the Metal That Builds Every Tank, Tool, and Transistor
Veron Wickramasinghe




I. THE DAY NOBODY NOTICED
On February 4, 2025, China’s Ministry of Commerce and the General Administration of Customs jointly published Announcement No. 10. The notice ran to a few hundred Chinese characters. It listed 41 Harmonized System codes covering tungsten, tellurium, bismuth, molybdenum, and indium. It required exporters to apply for licenses from MOFCOM and to prove that the buyer was not a military end-user. It took effect immediately.
It was published in the late morning, Beijing time. Most of the financial press was still writing about gallium and germanium, the previous controls. The trade lawyers at Pillsbury and Covington and Baker McKenzie filed it under “dual-use framework.” The procurement officers in Washington and Brussels did not call emergency meetings. The European APT benchmark, assessed weekly by Fastmarkets, was sitting at roughly $342 per metric ton unit, in line with its long-term average of approximately $300 where it had quietly traded for most of the post-Cold War era.

That was the last normal day in the global tungsten market.

Thirteen months later, in late March 2026, the same Fastmarkets benchmark printed above $3,000 per metric ton unit. The increase from the pre-control baseline exceeded 775 percent. CNBC reported the move on March 31, citing Almonty Industries citing Fastmarkets. Bloomberg’s George Heppel of BMO Capital Markets, a 12-year commodities veteran, said in print on March 15 that he had never seen a market this tight outside of lithium in 2021, and that this one was worse because there was no pipeline of projects waiting to come online. Lewis Black, the chief executive of Almonty, told Bloomberg the same day that the industrial base was desperate for material and that nobody really knew where the price was going to settle because, for the first time in modern memory, the market was actually being allowed to determine the tungsten price.

The largest non-Chinese tungsten mine to come online in over a generation, Almonty’s Sangdong project in South Korea, formally commissioned on March 16, 2026 with 200 South Korean government officials and US embassy representatives present. It will produce roughly 2,300 tonnes per year at full Phase I ramp. Global non-Chinese consumption sits north of 29,000 tonnes. The single most important Western tungsten event in 30 years closes less than 10 percent of the structural deficit.

The United States has not produced a single tonne of tungsten from a domestic mine since 2015. It imports approximately 27 percent of its tungsten supply directly from China. Its Defense Logistics Agency stockpile of tungsten alloy, according to GAO-24-106959, fell 100 percent between fiscal years 2019 and 2023. The DFARS regulation that bans Chinese tungsten from defense applications kicks into its full mine-to-product traceability phase on January 1, 2027. That is roughly 36 weeks from now.

This is the story of how the world sleepwalked into a tungsten crunch while watching the wrong commodities.

This is The Tungsten Throttle.

II. ONE NUMBER THAT TELLS THE STORY
Start with one number, because one number tells the entire story.
In February 2026, BMO Capital Markets analysts George Heppel and Helen Amos published a research note that crossed the wires of Mining.com, the Northern Miner, and a half-dozen other industry outlets. The note observed, in passing, that European APT prices had broken out of their long-term average of approximately $300 per tonne in 2025 and were trading at around $1,775 per tonne. By March 13, the trailing 12-month average had reached $2,250. By the last week of March, the spot benchmark was above $3,000 and the China Tungsten Industry Association was reporting European APT in a band of $2,800 to $3,150, up 223 percent year to date.

That move is the signal. A specialty industrial metal that nobody on financial television covers, with no futures contract, no ETF, and no household-name producers, has rerated by almost an order of magnitude in 13 months. The market knows something the procurement officers do not. The procurement officers were trained on the USGS Critical Minerals List and the European Critical Raw Materials Act. Both lists name tungsten. Neither list has translated naming into action at any scale that matters.

Now scale the signal globally. According to the USGS Mineral Commodity Summaries 2025, world mine production in 2024 was approximately 81,000 tonnes. China produced approximately 67,000 of those tonnes. That is roughly 83 percent of global supply from a single country, with the next largest producer, Vietnam, contributing 3,400 tonnes, and the United States contributing zero. China holds approximately 52 percent of global reserves. It also operates the only fully integrated downstream processing complex of meaningful scale, anchored by Xiamen Tungsten and the Ganzhou cluster in Jiangxi province.

That concentration is the headline number. It is also the wrong number to focus on alone, because raw production share understates how thoroughly China controls the chokepoint.

The number that determines what happens next is the export collapse. According to Shanghai Metals Market data citing China customs, total tungsten exports in January and February 2026 fell 27.6 percent year over year to roughly 1,800 tonnes. Exports of ammonium paratungstate, the form most directly used to feed Western downstream processors, collapsed to zero in late 2025 and stayed at zero through the first two months of 2026. BMO confirmed this in writing. Fastmarkets confirmed it in writing. The full-year 2025 APT export number fell roughly 70 percent against 2024.

China is not throttling tungsten by accident. China is throttling tungsten by policy, and the policy architecture, a state trading whitelist of 15 approved exporters published December 26, 2025, is structural rather than tactical. Unlike the rare earth controls that were partially suspended after the November 2025 US-China summit, the tungsten whitelist remains fully in force. As the Pillsbury Law analysis confirmed, tungsten was never inside the truce.

Three of the five largest non-Chinese consumers, Plansee in Austria, Sandvik in Sweden, and Mitsubishi Materials in Japan, depend on a Chinese feedstock supply that has effectively gone to zero for unapproved buyers. The geographic diversification of tungsten consumption is real. The geographic diversification of tungsten supply is theatre. Every road, until Sangdong’s ramp, ran through Ganzhou.

There is no strategic tungsten reserve in any Western country at meaningful scale. The US DLA stockpile is depleted and only now being rebuilt, with a planned acquisition of 2,041 tonnes of tungsten metal in the FY2025 Annual Materials Plan. There is no equivalent of the Strategic Petroleum Reserve. Tungsten was treated as a specialty industrial input for so long that the institutional infrastructure of scarcity around it was never built.

The world is now learning, in real time and at high speed, what happens when a metal that sits in everything from a Patriot interceptor to a Boeing landing gear to a Samsung NAND chip turns out to be controlled by one country with a national security framework around its export.

III. WHAT WALL STREET IS MISSING
Read the wires. Read the analyst notes. Watch financial television. The mainstream consensus on critical minerals through the first quarter of 2026 has organized itself around three commodities: rare earths, lithium, and copper.
Rare earths take the headlines because the 2010-2011 Molycorp story is still the reference parable in every commodities desk in New York and London. Every analyst over 40 remembers the dysprosium chart. Every analyst under 40 has read about it. When China tightens any specialty metal, the reflex is to map it onto the rare earth template.

Lithium takes the secondary attention because of the EV cycle and because Bloomberg can put a battery on the cover. Copper takes the tertiary because every electrification thesis runs through it.

The consensus narrative is therefore: this is a critical minerals story, but the action is in the metals you already know.

That consensus is incomplete to the point of being structurally wrong.

The consensus misses that tungsten is not really a critical mineral the way Wall Street normally thinks about critical minerals. It does not move EV unit economics. It does not anchor a battery thesis. It has no ETF of its own. It has no liquid futures contract. The largest pure-play producer outside China, Almonty Industries, had a market capitalization of roughly $50 million in early 2024 and roughly $4.8 billion by April 2026, having moved from $1 to a 52-week high of $22.55. Almost nobody on a Manhattan trading floor owned it on the way up.

Tungsten sits in cemented carbide cutting tools, in kinetic energy penetrators, in tungsten hexafluoride for semiconductor fabs, in jet engine counterweights, in radiation shielding for medical isotopes. It is four layers upstream of finished goods. The further upstream you move, the smaller the audience that understands what they are looking at.

Wall Street is staffed with people who can model a Tesla margin or a Freeport copper unit cost. Wall Street is not staffed with people who understand that the tungsten carbide insert on the lathe that machined the housing of the Patriot interceptor that the United States Navy fired in defense of Al Udeid Air Base on the night of the Iran strikes was almost certainly produced from Chinese APT, that the APT export to American buyers was zeroed in late 2025, and that the replacement supply chain runs through a single mine in South Korea that just commissioned six weeks ago.

There is a small group of people who do understand this. They are mostly metals industry veterans, defense industrial base analysts, and the named research desks at BMO, Fastmarkets, Project Blue, and the Modern War Institute at West Point. George Heppel and Helen Amos at BMO. Janine Le Roux at Project Blue, who told Bloomberg that military tungsten consumption including helicopters, fighter jets, and ammunition was set to grow 12 percent in 2026. The Fastmarkets editorial desk, which assessed in February 2026 that tungsten was now behaving like a strategic resource rather than a conventional industrial metal. None of these voices have been amplified by mainstream financial press in a way that an institutional generalist would actually notice.

That is the gap this article exists to close.

The consensus is treating tungsten as a footnote inside the critical minerals story. It is not a footnote. It is the spine of the cutting-tool industry, the spine of NATO ammunition, and the spine of legacy semiconductor manufacturing, all of which are now competing for a supply that has been cut at the regulatory source.

IV. THE GRADE TRAP

Here is the flaw at the heart of the consensus narrative. It is the single most important concept in this entire piece, and once you see it, you cannot unsee it.

The consensus assumes that if China relents, the supply problem unwinds. The consensus is wrong. China cannot relent enough to unwind the problem, because China’s own domestic supply is structurally tightening at the same time the export controls are biting.

Average ore grade in Chinese tungsten mines has declined from approximately 0.42 percent WO3 in 2004 to approximately 0.27 percent in 2024. That is roughly a 35 percent decline in head grade over two decades. The China Tungsten Industry Association reports that mining costs have breached 100,000 yuan per tonne of concentrate, a level at which marginal small and medium producers have voluntarily suspended production because they cannot make the economics work. Domestic mine operating rates fell below 35 percent by mid-2025.

China’s Ministry of Natural Resources first-batch mining quota for 2025 was set at 58,000 tonnes of 65 percent WO3 concentrate, a 6.45 percent cut from 2024. Jiangxi province, which produces roughly 60 percent of Chinese tungsten, took a 10.01 percent cut. Several provinces received a zero allocation. This is not a country with an abundant resource being strategic about exports. This is a country whose own mines are getting harder to work, whose costs are rising, and whose government is rationing the remaining production downward each year.

This is the grade trap. China weaponized tungsten exports at exactly the moment that even an unrestricted China would have been a tighter supplier than it was a decade ago. The export controls are the visible pressure. The grade decline is the invisible pressure underneath.

Now combine this with the demand side. CICC, the Chinese investment bank, forecasts that from 2026 to 2028 the global supply-demand gap will exceed 17 percent of demand, with a projected deficit of approximately 20,000 tonnes per year. Independent estimates from Jat Carbide and MeetYouCarbide put the figure between 19,000 to 21,000 tonnes annually. Western market participants describe a shortfall of roughly 13,000 tonnes of contained tungsten in non-Chinese markets specifically.

These numbers are not pricing in a world where China lifts its controls. These numbers are pricing in a world where the existing controls remain and Chinese ore grades continue their secular decline. If Beijing tomorrow rescinded Announcement No. 10 and reopened APT exports to all comers, the gap would narrow but it would not close. The structural deficit is bigger than any plausible policy reversal.

That is the most underappreciated fact in the entire crisis, and it is the reason this is a regime shift rather than a cycle.

V. THE PRICE ACTION
The price action tells the story. The price action does not lie.
The Fastmarkets European APT benchmark sat at $342 per metric ton unit in early February 2025. By the first trading session of 2026, it had reached 900 to $940. By January 29, 1,125 to 1,150. By mid-February, 1,650 to 1,900. By March 13, the trailing 12-month average had reached 2,250. By March 16, Bloomberg was reporting the spot price above 2,250 and noting that this represented a 557 percent increase since the export controls were imposed in February 2025.

By the last week of March, the spot benchmark was above $3,000. CNBC reported on March 31 that tungsten had hit a record high of over $3,000 late the previous week, marking a surge of over 50 percent for the month and more than tripling in price since late December.

Chinese domestic concentrate prices traveled the same arc. RMB 143,000 per tonne at the end of 2024 to RMB 1,005,000 per tonne by March 30, 2026. A 603 percent increase. Tungsten powder broke through RMB 2,000 per kilogram in the first quarter of 2026, up from RMB 316 at the end of 2024, a 533 percent increase.

Tungsten hexafluoride, the input that goes directly into semiconductor fab tungsten contact deposition, saw its export volume fall 38.9 percent year over year in January and February 2026 according to SMM. SK Specialty in South Korea and Japanese suppliers announced WF6 price hikes of 70 to 90 percent effective 2026. TrendForce noted in April 2026 that Chinese suppliers now hold over 50 percent of global WF6 market share, that semiconductor fabs consume 87 percent of global WF6 production, and that a potential WF6 supply halt from Japanese vendors could severely impact Samsung 3D NAND and DRAM production.

Then the corporate response confirmed the squeeze.

Kennametal, the most transparent listed pure-play on tungsten exposure, reported its fiscal second quarter on February 4, 2026. Chief executive Sanjay Chowbey told analysts that quarterly sales were driven by buy-ahead in response to the tungsten pricing environment. Sales reached $530 million, up 10 percent organic. Earnings per share rose 92 percent year over year. The company suspended share buybacks because working capital was tied up in higher-priced tungsten inventory. Kennametal was passing the cost through via surcharges, and customers were paying because there was no alternative supplier.

Almonty Industries, the only Western tungsten miner of meaningful scale to commission new production this cycle, saw its share price rise from $1 in early 2024 to a 52-week high of $22.55. Its market capitalization reached approximately $4.8 billion. B. Riley set a price target of $23. Plansee Group, the Austrian processor that holds a 15-year offtake on over 90 percent of Sangdong’s Phase I production, is also Almonty’s largest shareholder at 14 percent, the cleanest expression in any public market of a Western anchor buyer locking in non-Chinese supply.

EQ Resources on the Australian Securities Exchange returned 695 percent over 12 months. Fireweed Metals, holder of the Mactung deposit in the Yukon, the largest known high-grade tungsten resource in the Western world, took strategic investment from JX Advanced Metals of Japan in March 2026 in a C$61.5 million placement.

S&P Global, Bloomberg, CNBC, and Mining.com have all written extensively about the tungsten move in the last six weeks. None of these stories has reached the front page of the Wall Street Journal, the lead block of CNBC’s morning programming, or the cover of Bloomberg Businessweek. The signal is in the trade press. The consensus has not yet caught up.

This is not a forecast. This is what is already happening.

VI. THE CHEMISTRY OF THE CHOKEPOINT
Follow the metal. The chain is short, the metallurgy is well understood, and there is no substitute at scale for most applications.
Tungsten leaves a Chinese mine as concentrate, typically 65 percent WO3. The concentrate is processed into ammonium paratungstate, the workhorse intermediate from which almost every downstream tungsten product is made. APT is converted to tungsten oxide and then to tungsten metal powder via hydrogen reduction. The metal powder is then either pressed into mill products such as wire, rod, and sheet, or combined with carbon to produce tungsten carbide powder for cemented carbide. A separate fluorination route produces tungsten hexafluoride for semiconductor deposition.

This chemistry has been the dominant route for producing usable tungsten for over half a century. It is not complicated. It does require APT as the gateway intermediate, and it does require a country with the integrated downstream capacity to make it economic. China is that country. Plansee in Austria, Wolfram Bergbau und Hütten also in Austria and now part of Sandvik, H.C. Starck now inside Mitsubishi Materials, and a handful of Korean and Japanese processors are the entire non-Chinese mid-stream universe at any meaningful scale.

Once you have tungsten metal or tungsten carbide powder, you have entered the foundational metallurgy of industrial civilization. Approximately 55 to 60 percent of global tungsten consumption goes into cemented carbide. The USGS confirms in its Mineral Commodity Summaries that an estimated 60 percent of US tungsten consumption is used in cemented carbide parts. Cutting tools account for roughly half of carbide applications. Mining drill bits, oil and gas drilling equipment, and tunnel boring machines account for most of the rest. The global tungsten carbide market was valued at between $18 to $24 billion in 2023 to 2024.

In the defense sector, tungsten goes into kinetic energy penetrators, missile warheads, close-in weapon system ammunition, aircraft counterweights, and radiation shielding. The AGM-88C HARM missile warhead contains 12,845 tungsten alloy cubes. The US Navy Phalanx CIWS switched to tungsten penetrators in the late 1990s. Most NATO nations outside the United States, including Germany and most Leopard 2 operators, use tungsten rather than depleted uranium for kinetic penetrators on environmental and political grounds. Rheinmetall’s AHEAD ammunition system uses 152 tungsten sub-projectiles per 35mm round in older variants and 675 sub-projectiles per round in the newer PMD428 anti-drone variant.

In semiconductors, tungsten is deposited via WF6 chemical vapor deposition into the contact layers and word lines of every legacy and mainstream logic and memory chip in production today. Lam Research maintains over 1,000 ALTUS tungsten deposition modules in production globally, with 15 years of market leadership. Applied Materials introduced its Centris Spectral Molybdenum ALD system at SEMICON Korea in February 2026, replacing tungsten with molybdenum at the leading edge for 15 percent contact resistance reduction. Samsung is migrating from tungsten to molybdenum for 3D NAND word lines from V5 onward. Intel switched from tungsten to cobalt for contacts at 10 nanometer in 2017 and 2018. These transitions are real. They will take 5 to 10 years to displace meaningful tungsten demand. Legacy and mature nodes, which represent the vast majority of global fab capacity, continue to use tungsten and will continue to use it for the rest of this decade.

The USGS states the substitution problem with characteristic bluntness in its Mineral Commodity Summaries: tungsten substitution is limited and any substitutes typically result in increased cost or lower performance.

Fastmarkets put it in writing in February 2026: substitution remains nearly impossible for most industrial uses, and recycling cannot scale sufficiently in the short term. New mine developments in Europe, the United States, and Asia remain years away from entering supply chains.

Read that paragraph again. Now combine these two demand streams. Defense demand is rising, projected by Project Blue at 12 percent in 2026. Industrial cemented carbide demand is rising with global capital expenditure. Semiconductor WF6 demand is steady on legacy nodes. PV diamond wire tungsten demand has reached over 60 percent penetration in 2026 with annual incremental demand exceeding 4,500 tonnes. Four demand streams pulling on a supply that has been regulatorily severed.

The chemistry of the cemented carbide industry was built on Chinese APT. The chemistry of NATO ammunition was built on Chinese APT. The chemistry of legacy semiconductor manufacturing was built on Chinese APT. The chemistry has not changed. The supply has.

VII. THE GHOST OF MOLYCORP

There is exactly one historical parallel that captures what is happening to the tungsten market right now, and it is one of the most instructive industrial shocks of the modern era.

In September 2010, after a Chinese fishing trawler collided with Japanese coast guard vessels near the Senkaku Islands, China imposed a de facto embargo on rare earth element exports to Japan. The price of dysprosium oxide, which had traded at $91 per kilogram in January 2009, surged to $2,377 per kilogram by August 2011. A 2,500 percent increase. Lanthanum oxide moved from $13 per kilogram to $208. Cerium oxide spiked roughly 50 to 100 times.

Then it collapsed. Dysprosium oxide fell to approximately $180 per kilogram by 2016, a 92.4 percent decline from peak. Molycorp, which had IPO’d at $14 and peaked at $79.16 on the Mountain Pass restart story, filed for Chapter 11 bankruptcy in June 2015 with $1.7 billion in debt and a stock price of 35 cents. Lynas, the Australian producer that nearly went bankrupt during the collapse, survived and is today the largest non-Chinese rare earth producer in the world.

What broke the rare earth squeeze was a combination of four factors. Japanese demand fell 31 percent from 2011 to 2012 through substitution and thrifting in magnet design. New supply came online from Mountain Pass and Mount Weld. The WTO ruled China’s export quotas illegal, and China dropped them, with only 49 percent of the 2011 quota actually used. Speculative inventory hoarding unwound.

The tungsten 2026 setup is structurally different from the rare earth 2011 setup, and the differences matter for how this resolves.

Tungsten substitution is substantially harder than rare earth substitution. Magnet manufacturers could redesign to use less dysprosium and substitute neodymium-iron-boron with samarium-cobalt for some applications. There is no equivalent option for kinetic penetrators in non-DU NATO nations. There is no equivalent option for general-purpose carbide cutting tools. There is no equivalent option for high-temperature applications above 2,000 degrees Celsius. Tungsten has the highest melting point of any metal at 3,422 degrees Celsius. Molybdenum, the closest substitute, fails above approximately 1,500 degrees in oxidizing environments.

China’s tungsten controls are framed as national security dual-use measures under the October 2024 Regulations on Export Control of Dual-Use Items, not as export quotas. They are not vulnerable to a WTO challenge in the way the rare earth quotas were. Beijing has built a regulatory architecture around this control that did not exist in 2010.

China’s own domestic tungsten supply is tightening. This was not the case with rare earths, where Chinese resources were abundant and the constraint was purely policy. The grade trap means that even a full reversal of Announcement No. 10 would not restore 2023 export volumes.

Defense demand creates a structural price floor that did not exist for rare earths. The Iran war has demonstrated, in the most expensive possible way, that twenty-first century warfare consumes tungsten at industrial volumes. The Foreign Policy Research Institute documented 5,197 munitions across 30-five types fired in the first 96 hours, with replacement costs of $10 to $16 billion. By day 16, RUSI estimated 11,294 munitions fired at approximately $26 billion in replacement cost. Nine hundred and 40-three Patriot interceptors were used, 18 months of production. Three hundred and 70-five Tomahawk cruise missiles, 53 months of production at the current rate of 85 per year.

There are fewer viable non-China tungsten projects in the development pipeline than there were rare earth projects in 2010. Almonty’s Sangdong is the only meaningful new Western mine commissioning this cycle. Fireweed’s Mactung will not produce before 2030 at the earliest. Pilot Mountain in Nevada is 5-plus years from production with a $6.2 million dollar DPA Title III pre-feasibility grant from July 2025.

The price move so far, roughly five to ten times depending on which baseline you use, is smaller than the rare earth peak move of twenty-five times. There is less speculative froth to unwind. That suggests less violent mean reversion.

Some correction from the late March peak is plausible. Industrial demand destruction is already occurring. Recycling can incrementally contribute. A trade deal could partially reopen Chinese supply. The cobalt parallel, with peak-to-trough declines of 74 percent in 12 to 18 months across two cycles in the 2010s and 2020s, demonstrates that commodity spikes can reverse violently even with structural narratives.

But a full Molycorp-style 90 percent collapse is unlikely. David Argyle, co-founder of Arlington Innovation Partners, told Bloomberg in March that he expected a maximum 24-month window of the squeeze being frustrating and annoying. That is the realistic resolution path. A correction of 30 to 50 percent from the late March peak is plausible over 12 to 24 months. A return to $300 APT is not.

The lesson of Molycorp for tungsten investors is not that the bull case is wrong. The lesson is that the companies that survive a commodity spike are the ones with genuine cost advantages and realistic timelines, not promotional narratives. Almonty has a 15-year offtake at a $235 floor with Plansee, a 3-year defense offtake with Tungsten Parts Wyoming for Pennsylvania munitions production, and a producing mine in Portugal generating cash flow today. EQ Resources has two operating mines and a 5-year traxys offtake. Sandvik has vertical integration through Wolfram Bergbau and a recycling rate above 70 percent of used inserts. Kennametal has demonstrated pricing power via surcharges. These are the companies built to survive the journey, regardless of where the spot price settles in 2027.

VIII. THREE STREAMS, ONE METAL

The immediate effects fall in three streams. Each stream is a separate industry that has been quietly and simultaneously squeezed by the same chokepoint, through the same metal, in the same 13 months.

The Defense Stream

The Iran war exposed the prelogistical reality of NATO ammunition. The Modern War Institute at West Point, the Payne Institute, and the Foreign Policy Research Institute have all published analyses showing that the consumption rates of the first weeks of conflict cannot be sustained against the current production rates of the Western defense industrial base. Patriot interceptors were used at a rate that consumed 18 months of production. Tomahawks at 53 months. The Pentagon’s FY2027 budget request includes a 150 percent increase across munitions procurement accounts and Iran-specific munitions at $22 billion, up from $4.5 billion in FY2026, a 387 percent increase.

Every one of those munitions contains tungsten. Every Patriot fragmentation warhead, every Tomahawk shaped charge component, every Phalanx round, every AHEAD sub-projectile. Rheinmetall’s recent low triple-digit million euro AHEAD order from a European customer implies tungsten demand in the hundreds of tonnes for a single contract.

The DFARS regulation 252.225-7052 bans Chinese tungsten from defense applications. Through December 31, 2026, the ban applies to melting and production. From January 1, 2027, it extends to mining, refining, separation, melting, and production. Full mine-to-product traceability. General Dynamics OTS, which produces the 120mm KE-W A1 APFSDS-T tungsten penetrator round and the M1028 canister with tungsten ball projectiles, is the most directly exposed prime. Lockheed Martin, RTX, and General Dynamics all file Conflict Minerals disclosures confirming that tungsten is present in substantially all products.

Representative John Moolenaar, chairman of the House Select Committee on the Chinese Communist Party, said in June 2025 that tungsten dependency on China posed a significant risk to the US defense industrial base and warfighting readiness. Lewis Black of Almonty confirmed in the Bloomberg interview that US authorities had directly contacted his company seeking immediate material availability and that nearly half of Sangdong’s South Korean output was earmarked for Pennsylvania for munitions production.

This is what a tungsten-constrained defense industrial base looks like. The President orders quadruple munitions production. The contractors agree. The contracts are signed. And the carbide cutting inserts that machine the missile housings are now 33 percent more expensive year to date, the APT feedstock that makes those inserts is on a Chinese export ban, and the only meaningful new Western supply mine commissioned six weeks ago.

The Industrial Stream

Cemented carbide is 55 to 60 percent of global tungsten demand. Sandvik is the world’s largest cutting tool company. Kennametal is the most transparent listed pure-play. ISCAR, owned by Berkshire Hathaway through IMC Group, is the world’s number two and includes Tungaloy, Japan’s first cemented carbide developer.

Sandvik’s vice president Dean Kangleas told industry press that the company does not have to worry about tungsten supply because of vertical integration through Wolfram Bergbau und Hütten, the only integrated tungsten smelting plant outside Asia and Russia, which operates the Mittersill scheelite mine in Austria, the largest scheelite deposit in Europe. Sandvik buys back 90 percent of used inserts and recycles them at 70 percent less energy and 64 percent less CO2. This integration is the model. It is also unique. Almost no other Western downstream consumer has this level of insulation.

Kennametal has demonstrated that cost pass-through via surcharges works in this environment, with EPS up 92 percent year over year in fiscal Q2 2026 despite working capital pressure. The customers are paying.

Chinese downstream tool manufacturers, including Zhuzhou Cemented Carbide, gain a structural competitive advantage from a domestic two-tier pricing market. Chinese domestic APT trades at RMB 1,480,000 per tonne. European APT trades at $2,800 to $3,15$0 per metric ton unit. The arbitrage favors Chinese-located manufacturing. MeetYouCarbide, the industry trade publication, describes a fierce collision pressuring small and medium-sized Western enterprises.

This is the slow-motion industrial competitiveness story underneath the headline price action. Western SME tool manufacturers cannot pass through the full cost increase. Chinese competitors do not need to pay it.

The Semiconductor Stream

Tungsten hexafluoride is consumed at approximately 2,500 metric tonnes annually in semiconductor fabs globally, with Asia-Pacific accounting for 72 percent of demand. South Korea, primarily Samsung memory, consumes roughly 35 percent. Taiwan, principally TSMC, uses over 900 tonnes. Chinese suppliers now hold over 50 percent of global WF6 market share according to TrendForce.

China’s WF6 exports fell 38.9 percent year over year in January and February 2026 according to SMM. SK Specialty in South Korea and Japanese suppliers announced WF6 price hikes of 70 to 90 percent for 2026. TrendForce warned in April 2026 that a potential WF6 supply halt from Japanese vendors could severely impact Samsung 3D NAND and DRAM production.

The leading-edge migration from tungsten to molybdenum at Samsung 3D NAND and to cobalt at Intel logic is real, and Applied Materials’ Centris Spectral Molybdenum ALD launch in February 2026 was the public signal that the equipment ecosystem has caught up with the design intent. But the migration window is 5 to 10 years for meaningful demand displacement. Lam Research’s installed base of over 1,000 ALTUS tungsten modules is not going anywhere by the end of this decade.

In the meantime, every legacy fab in Taiwan, Korea, China, the United States, and Europe continues to deposit tungsten contacts using WF6 sourced predominantly from Chinese-controlled supply chains, at prices that just rose 70 to 90 percent.

Three streams. One metal. Three industries simultaneously absorbing the same upstream shock at the same time, because the upstream supply chain was built on a Chinese feedstock that is no longer available at scale to non-approved buyers.

IX. THE INVISIBLE REDRAWING OF THE MAP

The first-order effects are visible in price action and corporate guidance. The second-order effects are where the strategic redrawing of the global industrial map happens, and where the thoughtful capital opportunity sits.

The first second-order consequence is the structural bifurcation of the global tungsten market into a Chinese-domestic price tier and an allied-Western price tier, with no clear mechanism to reunify them. The Chinese domestic APT price translated to dollars sits well below the European benchmark even in renminbi terms. The export control wedge is now permanent because the regulatory architecture is permanent. The DFARS January 2027 cliff makes the wedge bidirectional. American defense buyers cannot legally source from China. Chinese exporters cannot legally ship to American military end-users. Two markets, one metal, no convergence.

The second consequence is the geopolitical realignment around tungsten supply security. The United States Department of Defense awarded $15.8 million in DPA Title III funding to Fireweed Metals for Mactung in December 2024 and $6.2 million to Golden Metal Resources for Pilot Mountain in July 2025. Canada has put C$12.9 million into the Critical Minerals Infrastructure Fund for Mactung. Japan’s JX Advanced Metals invested C$61.5 million in Fireweed in March 2026. South Korea hosted the Sangdong commissioning with 200 government officials and US embassy presence. The five-eyes plus Japan plus Korea axis is being woven around tungsten projects in the same way it was woven around lithium and rare earths over the past five years, but more quietly.

The third consequence is the elevation of midstream processors as the strategic chokepoint. Plansee in Austria, Wolfram Bergbau und Hütten now inside Sandvik, H.C. Starck now inside Mitsubishi Materials, and Tungsten Parts Wyoming inside Israel’s Metal-Tech are the entire non-Chinese mid-stream universe at any meaningful scale. Whoever controls this layer controls the conversion of mine output into usable industrial product. Plansee’s 14 percent stake in Almonty and its 15-year Sangdong offtake are the cleanest expression of vertical integration as strategic insurance in any Western critical minerals market today. Watch for further consolidation in this layer over the next 18 months.

The fourth consequence is the redrawing of critical minerals policy itself. Tungsten is on the USGS Critical Minerals List. It is on the EU Critical and Strategic lists. It is on the UK 2024 Criticality Assessment. It is named in the UK Vision 2035 Critical Minerals Strategy. Australia has tungsten under the Critical Minerals Production Tax Incentive at 10 percent for 2027 to 2040. The lists exist. The funding is starting to flow. The question is whether the policy clock can outrun the depletion clock on the DLA stockpile and the procurement clock on the FY2027 munitions budget. The honest answer is that it cannot, in this cycle. The policy infrastructure was built 5 to 10 years too late.

The fifth consequence is the collapse of the geographic diversification thesis in tungsten downstream. Standard analyst notes will tell you that global tungsten consumption is geographically diverse, with major operations in Europe, North America, Japan, and Korea. This is true. It is also strategically irrelevant. Every major non-Chinese consumer depends on a feedstock supply chain that until 13 months ago ran through Ganzhou. The diversification of consumption was real. The diversification of supply was theatre. The crisis just made the theatre expensive.

X. THE HONEST COUNTER-CASE

There are honest counterarguments to this thesis, and any serious analyst is obligated to engage them seriously rather than dismiss them.

The first counterargument is that a US-China trade deal could partially reopen Chinese supply within weeks. This is true mechanically. The MOFCOM whitelist could be expanded. Individual export licenses could be issued. The honest response is that the policy architecture, the state trading whitelist and the dual-use control framework, was built specifically to be durable against political pressure. Tungsten was excluded from the November 2025 rare earth truce. Beijing has signaled by action that tungsten is a different category of control, more akin to a permanent strategic export regime than a tactical bargaining chip. A partial reopening is possible. A return to 2023 export volumes is not.

The second counterargument is that demand destruction is already occurring and will accelerate. IMARC reports US tungsten prices fell 13.3 percent quarter over quarter in Q4 2025 as hard metals fabricators and tooling producers cut order volumes sharply. European scrap prices softened in April 2026 as holders began destocking. PV industry production cuts of approximately 40 percent in operating rates temporarily reduced tungsten wire demand in mid-2025. The honest response is that defense demand is counter-cyclical and growing 12 percent per Project Blue, that legacy semiconductor demand is sticky, and that the 2008-2009 precedent shows industrial tungsten demand can fall 15 to 25 percent in a severe recession but is partially offset by rising defense and stockpile procurement. Demand destruction will moderate the spike. It will not break the structural deficit.

The third counterargument is that recycling can scale faster than the bull case allows. Current recycling provides approximately 30 to 35 percent of global tungsten supply per ITIA. Cutting insert recycling rates already reach above 95 percent in industrial settings. The honest response is that pushing the overall rate from 35 to 45 percent over 2 to 3 years could add 5,000 to 8,000 tonnes of supply, which is meaningful but insufficient to close a 20,000 tonne annual deficit. Recycling helps. Recycling does not solve.

The fourth counterargument is that Almonty’s Sangdong ramp could be faster than projected. Phase I commissioning in March 2026 is the easy part. Phase II would double output and is fully permitted. The honest response is that Almonty has a track record of delays. Sangdong was originally targeted for H2 2025. The Phase II expansion is unlikely to deliver meaningful additional tonnage before late 2027 at the earliest. Even at full Phase II run rate, Sangdong supplies roughly 40 percent of non-China demand. It does not close the gap alone.

The fifth counterargument is that semiconductor migration to molybdenum and cobalt will permanently dampen WF6 demand. This is true on a 5 to 10 year horizon. It is not true on the 24-month horizon that determines this price cycle. Lam Research’s installed base of over 1,000 tungsten ALTUS modules will continue to consume WF6 for the rest of the decade.

The strongest version of the bear case is that prices moderate from the late March peak as demand destruction, recycling, Sangdong ramp, and a partial trade deal combine to ease the squeeze. All of this is plausible. None of it changes the structural deficit. The price will likely correct from the $3,000 peak. It will not return to $300. The two-tier market will remain. The DFARS cliff will arrive. The defense procurement cycle will tighten.

The bear case is right about the destination. The bear case is wrong about the journey.

The journey is what creates the trade.

XI. THE WATCHLIST

Here is what to monitor over the next 90 days, in order of strategic importance. These are the indicators that will tell you whether the thesis is working in real time.

Monthly China customs tungsten export data via GACC, SMM, and CTIA is the primary signal. Any reversal from the 27.6 percent year-over-year decline indicates a policy shift. Continued decline confirms structural lockdown. Particular attention to APT-specific export volumes, which were zero in late 2025 and through early 2026.

Weekly Fastmarkets European APT benchmark is the single most important price indicator. Sustained prints below $2,000 per metric ton unit would suggest meaningful cooling. Sustained prints above $3,000 confirm the regime shift is intensifying.

MIIT and Ministry of Natural Resources second-batch 2026 mining quota announcements will signal Beijing’s supply intent for the second half of 2026. A further cut would confirm structural tightening. A surprise increase, while unlikely, would moderate the bull case.

Almonty Industries quarterly production updates from Sangdong are the key Western supply variable. The Phase I ramp curve in Q2 and Q3 2026 will determine how quickly non-Chinese supply can absorb defense procurement demand. Pay particular attention to any commentary on Phase II timing.

DLA stockpile acquisition tenders for tungsten are the institutional buyer signal. Large purchases would tighten spot markets and accelerate price discovery. The FY2025 Annual Materials Plan calls for 2,041 tonnes of tungsten metal acquisitions. The pace of execution matters.

DFARS waiver and extension decisions are the regulatory signal. Any Congressional action to delay the January 2027 full supply-chain ban beyond the already-granted 1-year extension would moderate the squeeze on defense primes. Absence of action confirms the cliff is binding.

Chinese domestic APT pricing versus international pricing is the bifurcation signal. Any narrowing of the two-tier price gap would suggest export flow resumption. Continued widening confirms the market is permanently fragmented.

Kennametal quarterly earnings are the cleanest listed-company signal on tungsten cost pass-through. Sales, margin, and surcharge commentary will reveal whether the industrial demand is holding up to the price increases.

Plansee, Sandvik, and Mitsubishi Materials commentary on tungsten procurement and recycling is the strategic mid-stream signal. Any major new offtake announcement, capacity expansion, or processing investment will reshape the non-Chinese supply chain.

Semiconductor industry molybdenum and cobalt adoption rates at TSMC, Samsung, and Intel are the secular demand signal. Accelerated transition from tungsten at the leading edge would create a long-term headwind, but the timing matters. Any 2026 announcements on new node tungsten replacement should be tracked.

If three or more of these indicators move in the direction the analysis predicts, the thesis is working. If three or more move against it, the bear case is winning and positioning should be reconsidered.

XII. THE GREY METAL

There is a grey metal you have never thought about.

It machines every cylinder block and every turbine blade in the modern industrial economy. It penetrates every kinetic round fired by a non-American NATO tank. It deposits every contact layer in every legacy semiconductor chip. It cuts every silicon wafer for every solar panel. It anchors every counterweight in every Boeing landing gear. It sits four layers upstream of almost every finished product in the manufacturing world. It is produced overwhelmingly in one country. That country has just placed it under national security export control.

There is no strategic reserve at meaningful scale in any Western country. There is no equivalent of the Strategic Petroleum Reserve. There is no liquid futures contract. There is no household-name producer outside of one Toronto-listed company that traded for $1 a share 2 years ago. The largest seaborne supplier is China. The second largest is Vietnam. The third largest is Russia. Almost all of the global mid-stream processing capacity sits in three Austrian and Japanese firms that depend on Chinese feedstock that is no longer flowing.

Thirteen months ago, Beijing weaponized that supply.

The European APT benchmark has moved from $342 per metric ton unit to over $3,000. The largest American defense ammunition prime is staring at a January 2027 regulatory cliff that requires full mine-to-product traceability for non-Chinese supply. The largest Western tungsten miner just commissioned its first non-Chinese mine in over a generation, and it closes less than 10 percent of the structural deficit. The semiconductor industry is paying 70 to 90 percent more for the WF6 that goes into every legacy node in production. The cemented carbide industry is watching Chinese competitors gain a structural cost advantage that did not exist 18 months ago. The defense procurement cycle is competing with industrial buyers for the same shrinking pool of non-Chinese supply.

None of this is hypothetical. All of it is happening right now, this week, while the financial press writes about lithium and rare earths and copper. The mainstream consensus has organized itself around the wrong critical minerals. The procurement officers are looking at the wrong list. The strategic stockpile that should have existed for tungsten was never built, because nobody bothered to define tungsten as urgent, because tungsten was a specialty industrial input, because a specialty industrial input is not strategic until the day it is.

Today is that day.

In 2010, when China embargoed rare earth exports to Japan, the global response was a 5-year scramble that ended in Molycorp’s bankruptcy and Lynas’s survival. The price spiked twenty-five times and then collapsed 90 percent. The lesson was supposed to be that the West would never again be caught flat-footed on a Chinese critical mineral chokepoint. The lesson was supposed to be that strategic stockpiles, supply diversification, and downstream processing investment would be made in advance of the next crisis.

Fifteen years later, the West has been caught flat-footed on a Chinese critical mineral chokepoint. The strategic stockpile is depleted. The supply diversification is one mine in South Korea that just commissioned. The downstream processing investment is a 14 percent stake by an Austrian processor in a Toronto-listed miner.

There is no Mountain Pass for tungsten. There is no Mount Weld. There is no Lynas. There is one Almonty, one Plansee, one Sandvik, one Kennametal, and a Mactung deposit in the Yukon that will not produce before 2030.

There is a grey metal you have never thought about.

You will think about it now.

Veron Wickramasinghe writes on geopolitics, energy, and the upstream supply chains that shape global power. The Tungsten Throttle is part of an ongoing series on the structural vulnerabilities of the modern industrial economy. Earlier pieces in the series include the Sulfur Trap, the Methane Trilemma, and the Qatar Helium analysis. Follow @veronken for the next installment.

Disclaimer

This piece is analysis, not investment advice. The companies, tickers, and price targets discussed are presented for analytical and informational purposes only. Readers should conduct independent due diligence and consult qualified advisors before making investment decisions. The author has no active position in the securities named at the time of publication and discloses any positions where taken in subsequent updates.

Sources

All quantitative claims in this piece are verified against primary or Tier 1 sources, including the United States Geological Survey Mineral Commodity Summaries 2025, Fastmarkets European APT benchmark assessments, Bloomberg reporting by James Attwood, Annie Lee, and Jacob Lorinc (March 15, 2026), CNBC reporting (March 31, 2026), BMO Global Commodities Research notes by George Heppel and Helen Amos (February 23 and March 2026), Project Blue analysis cited via Janine Le Roux, Reuters reporting on the MOFCOM whitelist (December 30, 2025), MOFCOM and General Administration of Customs Announcement No. 10 (February 4, 2025), the China Tungsten Industry Association, Shanghai Metals Market customs data, the Kennametal Q2 FY2026 earnings call transcript and SEC filings, Almonty Industries SEC filings and Sangdong commissioning announcements, GAO-24-106959 on DLA stockpile depletion, the DFARS 252.225-7052 regulation, the Modern War Institute at West Point, the Foreign Policy Research Institute, RUSI updated Iran conflict munitions analysis, the Atlantic Council, the Center for Strategic and International Studies, TrendForce WF6 market analysis (April 2026), Applied Materials press releases (February 10, 2026), and SemiAnalysis coverage of semiconductor metallization. Errors of interpretation are the author’s alone.